Programmes of education in breast self-examination with specialist clinics for self-referral were introduced in two health districts around 1980. Combining the results from the two centres showed no reduction in mortality from breast cancer over the following 10 years but the mortality was low in one of the centres whilst in the other it was higher than in four geographically separate comparison centres in which there was similar careful monitoring of breast cancer incidence and mortality. Because this was not a randomised controlled trial and lacked a uniform treatment protocol, biases may be responsible for the differences observed, but it is also possible that BSE education with annual reinforcement contributed to the breast cancer mortality reduction seen in one district. The overall conclusion however is that the value of breast self-examination remains unproven.